Gene identification in common disorders such as Alzheimer disease and breast cancer has greatly profited from the use of age of onset as criterion to delineate subgroups of disease characterized by different inheritance patterns. In bipolar affective disorder, where the majority of linkage studies have produced conflicting results, studies reporting clinical characteristics and familial occurrence of disease have suggested that age of onset might serve as an indicator for identifying more homogeneous subgroups of disease. Our study was the first to examine this hypothesis by the means of segregation analysis. We investigated a sample of 177 bipolar I probands recruited from consecutive admissions and their first- and second-degree relatives (2,407 subjects). Probands were subdivided into an early-onset (n = 107) and a late-onset group (n = 70) using an age of onset of 25 as a cut-off point. This age was chosen because the observed age of onset distribution was bimodal with a cut-off of 25 years. Morbid risks for affective disorder were found significantly higher (P = 0.01) in relatives of probands with an early onset than in probands with late onset of disease. The segregation analysis showed that the disease is transmitted differently in early- and late-onset groups. In the early-onset group, a non-Mendelian major gene with a polygenic component was favored while the data in the late-onset group were compatible with a multifactorial model. This result may have important implications for future molecular studies aiming at the identification of disease-associated genes.
Copyright 2001 Wiley-Liss, Inc.